fredmiranda.com
Login
This topic is in the 2024 archive. You can read it and like its posts here. It is read only.

Archive 2024A9III Thoughts from the first few hours

  
 
docusync
Offline
Details
p.2 #1A9III Thoughts from the first few hours
Has anybody submitted a $50 rebate claim for their A9III? 0.83% back!
https://rebate.eur.electronics.sony.com/#/ceur/product-registration?source=dtc_us
Feb 09, 2024 at 10:45 PM
Tony Ross
Offline
Details
p.2 #2A9III Thoughts from the first few hours
jhapeman wrote:
So my A9III arrived this morning and here's some quick thoughts for those wondering about it. Not really any photo opportunities, more about the changes in the camera ergonomics, button layout, etc.

1) Ergonomics: I've never been put off by the small size of the Sony cameras, and have in fact felt it was an advantage. However, I have felt there was room for improvement, especially with the cramped space between the grip and lens mount. That is very nice now, plenty of room. The new grip area around shutter is a big improvement, and they made some button changes there
...Show more →

3) - no, I believe the new battery grip is NOT backward compatible. It's quite a bit more expensive, too, but they made a big thing about all the buttons and switch on the grip being the same quality as the ones on the camera - gess they must have been using lower quality ones in the past? The new grip also ekes a bit more power from a pair of batteries, which is interesting.

4) - as far as I can tell the rear screen is identical to the one of the A7RV, and I've been enjoying that one for over a year. The EVF is the same, too, but has the bonus of supporting 120fps without shrink.

5) - yes, there are a bunch more settings in the menus. Some are necessary: things like supporting the Turbo button (setting its speed), and controlling the composite RAW. And there are some extras like multiple sets of something or other (I gather that stuff matters tones organisation!) Not to mention the settings allowing the setup of non-Sony flashes for sync at high shutter speeds.

6) - and there's the XL spot as well as the XS.

7) - I'm hoping it also still supports updates from card. I really don't want to install that Creator App thing.

Yes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)

Feb 10, 2024 at 04:17 AM
Tony Ross
Offline
Details
p.2 #3A9III Thoughts from the first few hours
randomguy wrote:
A small thing that I don't like in the A93 is that only the hi+ setting can be set to 60 or 120FPS.

I can make up to 4 different "boost" buttons with different FPS from 5 to 120 FPS but I can't set H to 60FPS for instance.


You are right: Lo / Mid / Hi can be set to 5 / 10 / 15 / 20 / 30. Only Hi+ can be set to 60 and 120 (oh, and the speed boost button, too).

That's odd - I could have sworn that the default for Hi was 60.
Feb 10, 2024 at 04:50 AM
randomguy
Offline
Details
p.2 #4A9III Thoughts from the first few hours
Tony Ross wrote:
randomguy wroteA small thing that I don't like in the A93 is that only the hi+ setting can be set to 60 or 120FPS.

I can make up to 4 different "boost" buttons with different FPS from 5 to 120 FPS but I can't set H to 60FPS for instance.
You are right: Lo / Mid / Hi can be set to 5 / 10 / 15 / 20 / 30. Only Hi+ can be set to 60 and 120 (oh, and the speed boost button, too).

That's odd - I could have sworn that the default for Hi was 60.


Yeah it is odd, but at least we can set boost to 5FPS .

I am sure they could fix it in firmware if it was possible to give them feedback somehow. I would just like to be able to jump between 60 and 120 without having to hold a button, as the buffer lasts twice as long with 60.
Feb 10, 2024 at 07:34 AM
j4nu
Offline
Details
p.2 #5A9III Thoughts from the first few hours
Tony Ross wrote:
4) - as far as I can tell the rear screen is identical to the one of the A7RV, and I've been enjoying that one for over a year. The EVF is the same, too, but has the bonus of supporting 120fps without shrink.


Does A7RV EVF shrink at 120fps? It's 240fps for A1...
Feb 10, 2024 at 07:35 AM
randomguy
Offline
Details
p.2 #6A9III Thoughts from the first few hours
Tony Ross wrote:
Yes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)



Yes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.
Feb 10, 2024 at 07:51 AM
snapsy
Offline
Details
p.2 #7A9III Thoughts from the first few hours
randomguy wrote:
Tony Ross wroteYes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)
Yes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.


Can you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?
Feb 10, 2024 at 08:09 AM
randomguy
Offline
Details
p.2 #8A9III Thoughts from the first few hours
snapsy wrote:
randomguy wrote
Tony Ross wroteYes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)
Yes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.
Can you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?


You can obviously fit a larger number of compressed files in a fixed amount of storage, than the larger uncompressed files. The lossless compressed files are not much larger than the lossy compressed files so you would expect a similar number of files in the buffer. But that is not what is happening. So my suspicion is that lossy compressed files are compressed before they are put in buffer while the lossless are not. Perhaps already on the sensor.
Feb 10, 2024 at 08:21 AM
snapsy
Offline
Details
p.2 #9A9III Thoughts from the first few hours
randomguy wrote:
snapsy wrote
randomguy wrote
Tony Ross wroteYes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)
Yes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.
Can you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?
You can obviously fit a larger number of compressed files in a fixed amount of storage, than the larger uncompressed files. The lossless compressed files are not much larger than the lossy compressed files so you would expect a similar number of files in the buffer. But that is not what is happening. So my suspicion is that lossy compressed files are compressed before they are put in buffer while the lossless are not. Perhaps already on the sensor.


For context here are the numbers from the A9 III manual:

Buffer Capacity for continuous shooting:
raw uncompressed: 96 images
raw lossless compressed: 96 images
raw lossy compressed: 192 images

Number of files that fit on a 64GB card:
raw uncompressed: 1100
raw lossless compressed: 1800
raw lossy compressed: 2000

Like previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source). The 12-bit raw frames in memory will take up less space than 14-bit frames even before compression is applied. I'm guessing the buffer capacity relates to that, and perhaps also due to the performance difference of the compression algorithm, and that both lossy and lossless compression are compressed at the same stage in the processing pipeline.
Feb 10, 2024 at 08:57 AM
randomguy
Offline
Details
p.2 #10A9III Thoughts from the first few hours
snapsy wrote:
randomguy wrote
snapsy wrote
randomguy wrote
Tony Ross wroteYes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)
Yes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.
Can you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?
You can obviously fit a larger number of compressed files in a fixed amount of storage, than the larger uncompressed files. The lossless compressed files are not much larger than the lossy compressed files so you would expect a similar number of files in the buffer. But that is not what is happening. So my suspicion is that lossy compressed files are compressed before they are put in buffer while the lossless are not. Perhaps already on the sensor.
For context here are the numbers from the A9 III manual:

Buffer Capacity for continuous shooting:
raw uncompressed: 96 images
raw lossless compressed: 96 images
raw lossy compressed: 192 images

Number of files that fit on a 64GB card:
raw uncompressed: 1100
raw lossless compressed: 1800
raw lossy compressed: 2000

Like previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source). The 12-bit raw frames in memory will take up less space than 14-bit frames even before compression is applied. I'm guessing the buffer capacity relates to that, and perhaps also due to the performance difference of
...Show more →

12vs14 does not explain a doubling of the number of files in the buffer. 8bit per pixel as in lossy compressed file vs 16bit (uncompressed) does.

The number of images that fit on a card seem inaccurate to me, a sample ISO 3200 image out of my window just now is 27.3Mb lossy vs 31.9 lossless.
Feb 10, 2024 at 09:06 AM
snapsy
Offline
Details
p.2 #11A9III Thoughts from the first few hours
snapsy wrote:
randomguy wrote
snapsy wrote
randomguy wrote
Tony Ross wroteYes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)
Yes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.
Can you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?
You can obviously fit a larger number of compressed files in a fixed amount of storage, than the larger uncompressed files. The lossless compressed files are not much larger than the lossy compressed files so you would expect a similar number of files in the buffer. But that is not what is happening. So my suspicion is that lossy compressed files are compressed before they are put in buffer while the lossless are not. Perhaps already on the sensor.
For context here are the numbers from the A9 III manual:

Buffer Capacity for continuous shooting:
raw uncompressed: 96 images
raw lossless compressed: 96 images
raw lossy compressed: 192 images

Number of files that fit on a 64GB card:
raw uncompressed: 1100
raw lossless compressed: 1800
raw lossy compressed: 2000

Like previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source). The 12-bit raw frames in memory will take up less space than 14-bit frames even before compression is applied. I'm guessing the buffer capacity relates to that, and perhaps also due to the performance difference of
...Show more →

Another possibility is that Sony's BionZ image processing pipeline can natively work on lossy compressed data but not on lossless compressed data. This means that for lossless compression the image would have to be maintained as uncompressed until the full processing pipeline is done, whereas for lossy compression it could enter the pipeline as compressed. This is along the lines of what you were suspecting but not something that could be fixed with a firmware change. I kind of like this theory because both uncompressed and lossless compression seemed to have been added as an afterthought by Sony [due to customer complaints].
Feb 10, 2024 at 09:09 AM
snapsy
Offline
Details
p.2 #12A9III Thoughts from the first few hours
randomguy wrote:
snapsy wrote
randomguy wrote
snapsy wrote
randomguy wrote
Tony Ross wroteYes, ACR supports lossless compressed RAW-L and lossy compressed RAW from the A9 III (I tested the lossy compressed RAW unintentionally - it's the default RAW format.)
Yes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.
Can you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?
You can obviously fit a larger number of compressed files in a fixed amount of storage, than the larger uncompressed files. The lossless compressed files are not much larger than the lossy compressed files so you would expect a similar number of files in the buffer. But that is not what is happening. So my suspicion is that lossy compressed files are compressed before they are put in buffer while the lossless are not. Perhaps already on the sensor.
For context here are the numbers from the A9 III manual:

Buffer Capacity for continuous shooting:
raw uncompressed: 96 images
raw lossless compressed: 96 images
raw lossy compressed: 192 images

Number of files that fit on a 64GB card:
raw uncompressed: 1100
raw lossless compressed: 1800
raw lossy compressed: 2000

Like previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source). The 12-bit raw frames in memory will take up less space than 14-bit frames even before compression is applied. I'm guessing the buffer capacity relates to that, and perhaps also due to the performance difference of the compression algorithm, and that both lossy and lossless compression are compressed at the same stage in the processing pipeline.
12vs14 does not explain a doubling of the number of files in the buffer. 8bit per pixel as in lossy compressed file vs 16bit (uncompressed) does.

The number of images that fit on a card seem inaccurate to me, a sample ISO 3200 image out of my window just now is 27.3Mb lossy vs 31.9 lossless.


Going by the Nikon Z6's 12 vs 14 bit lossless compression, the 12-bit is 15% smaller. And that's only accounting for the final output. It's unclear if the in-memory arrangement of data would match this.

The capacity will likely vary based on image content and noise. You're seeing a 17% reduction in size vs the 11% implied by Sony's sample table. Doesn't seem that far off to me when also accounting for the potential differences on assumed average scene content.
Feb 10, 2024 at 09:18 AM
randomguy
Offline
Details
p.2 #13A9III Thoughts from the first few hours
snapsy wrote:
Like previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source).


By the way this does not seem to be the case. I shot a black frame in a 120FPS burst, lossy compressed and rawdigger histogram show pixel values of 2, 4, 6 which is consistent with the 13bit limit of the lossy compression.
Feb 10, 2024 at 09:26 AM
randomguy
Offline
Details
p.2 #14A9III Thoughts from the first few hours
snapsy wrote:
randomguy wrote
snapsy wrote
randomguy wrote
snapsy wrote
randomguy wroteYes actually note that the buffer capacity is halved with lossless compressed RAW. I guess the firmware compresses the file when it is written to the card and not before buffering it. This is also something they should be able to fix with an update.
Can you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?
You can obviously fit a larger number of compressed files in a fixed amount of storage, than the larger uncompressed files. The lossless compressed files are not much larger than the lossy compressed files so you would expect a similar number of files in the buffer. But that is not what is happening. So my suspicion is that lossy compressed files are compressed before they are put in buffer while the lossless are not. Perhaps already on the sensor.
For context here are the numbers from the A9 III manual:

Buffer Capacity for continuous shooting:
raw uncompressed: 96 images
raw lossless compressed: 96 images
raw lossy compressed: 192 images

Number of files that fit on a 64GB card:
raw uncompressed: 1100
raw lossless compressed: 1800
raw lossy compressed: 2000

Like previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source). The 12-bit raw frames in memory will take up less space than 14-bit frames even before compression is applied. I'm guessing the buffer capacity relates to that, and perhaps also due to the performance difference of the compression algorithm, and that both lossy and lossless compression are compressed at the same stage in the processing pipeline.
12vs14 does not explain a doubling of the number of files in the buffer. 8bit per pixel as in lossy compressed file vs 16bit (uncompressed) does.

The number of images that fit on a card seem inaccurate to me, a sample ISO 3200 image out of my window just now is 27.3Mb lossy vs 31.9 lossless.
Going by the Nikon Z6's 12 vs 14 bit lossless compression, the 12-bit is 15% smaller. And that's only accounting for the final output. It's unclear if the in-memory arrangement of data would match this.

The capacity will likely vary based on image content and noise. You're seeing a 17% reduction in size vs the 11% implied by Sony's sample table. Doesn't seem that far off to me when also accounting for the potential differences on assumed average scene content.


I am seeing a much bigger reduction of size with lossless compression, a uncompressed raw has exactly 16bit per pixel while the lossy compressed has 8bit. Which should put the uncompressed files at 24Mb larger than lossy compressed or 51.3Mb vs 31.9 lossless. Which is almost 38% reduction.

Edited on Feb 10, 2024 at 09:33 AM · View previous versions
Feb 10, 2024 at 09:32 AM
snapsy
Offline
Details
p.2 #15A9III Thoughts from the first few hours
randomguy wrote:
snapsy wroteLike previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source).
By the way this does not seem to be the case. I shot a black frame in a 120FPS burst, lossy compressed and rawdigger histogram show pixel values of 2, 4, 6 which is consistent with the 13bit limit of the lossy compression.


Interesting. The extra bit resolution could be a side effect of something Sony added in the image processing chain. I would tend to give preference to Sony's documentation on this - they specifically updated the document for the A9M3.
Feb 10, 2024 at 09:32 AM
snapsy
Offline
Details
p.2 #16A9III Thoughts from the first few hours
randomguy wrote:
snapsy wrote
randomguy wrote
snapsy wrote
randomguy wrote
snapsy wroteCan you explain why you believe the timing of the image compression within the shooting/processing pipeline affects the buffer capacity?
You can obviously fit a larger number of compressed files in a fixed amount of storage, than the larger uncompressed files. The lossless compressed files are not much larger than the lossy compressed files so you would expect a similar number of files in the buffer. But that is not what is happening. So my suspicion is that lossy compressed files are compressed before they are put in buffer while the lossless are not. Perhaps already on the sensor.
For context here are the numbers from the A9 III manual:

Buffer Capacity for continuous shooting:
raw uncompressed: 96 images
raw lossless compressed: 96 images
raw lossy compressed: 192 images

Number of files that fit on a 64GB card:
raw uncompressed: 1100
raw lossless compressed: 1800
raw lossy compressed: 2000

Like previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source). The 12-bit raw frames in memory will take up less space than 14-bit frames even before compression is applied. I'm guessing the buffer capacity relates to that, and perhaps also due to the performance difference of the compression algorithm, and that both lossy and lossless compression are compressed at the same stage in the processing pipeline.
12vs14 does not explain a doubling of the number of files in the buffer. 8bit per pixel as in lossy compressed file vs 16bit (uncompressed) does.

The number of images that fit on a card seem inaccurate to me, a sample ISO 3200 image out of my window just now is 27.3Mb lossy vs 31.9 lossless.
Going by the Nikon Z6's 12 vs 14 bit lossless compression, the 12-bit is 15% smaller. And that's only accounting for the final output. It's unclear if the in-memory arrangement of data would match this.

The capacity will likely vary based on image content and noise. You're seeing a 17% reduction in size vs the 11% implied by Sony's sample table. Doesn't seem that far off to me when also accounting for the potential differences on assumed average scene content.
I am seeing a much bigger reduction of size with lossless compression, a uncompressed raw has exactly 16bit per pixel while the lossy compressed has 8bit. Which should put the uncompressed files at 24Mb larger than lossy compressed or 51.3Mb vs 31.9 lossless. Which is almost 38% reduction.


I was using the lossy vs lossless file sizes you posted (27.3Mb vs 31.9), which represents a 17% reduction.
Feb 10, 2024 at 09:35 AM
randomguy
Offline
Details
p.2 #17A9III Thoughts from the first few hours
snapsy wrote:
randomguy wrote
snapsy wroteLike previous Sony bodies, the A9 III switches from 14-bit to 12-bit readout when using lossy compression in continuous shooting mode (source).
By the way this does not seem to be the case. I shot a black frame in a 120FPS burst, lossy compressed and rawdigger histogram show pixel values of 2, 4, 6 which is consistent with the 13bit limit of the lossy compression.
Interesting. The extra bit resolution could be a side effect of something Sony added in the image processing chain. I would tend to give preference to Sony's documentation on this - they specifically updated the document for the A9M3.


That is possible, there is some weirdness in the distribution that might be from rounding actually.
Feb 10, 2024 at 09:41 AM
robert614
Offline
Details
p.2 #18A9III Thoughts from the first few hours
Just got mine yesterday. Haven’t had a chance to shoot anything yet. Just playing around with it and setting it up.

First impressions:

Grip is much better. Going back and forth between the A9III and the A1, the A1 feels so much more boxy and hard edged. The A9III’s grip is nicely rounded in the hand. And the weird way the fingertip grip area of the A1 angles away from the lens mount is gone.

Grip could still be taller IMO. Another 5-10mm would be perfect for me. With the new vertical grip attached though, the camera feels very nice in the hand. For more casual use, I would still like a grip extender. I hope someone comes out with one soon. Might take a while though. Since the A9III is a low volume product.

Overall though, the ergonomics are a big step in the right direction.

Buttons feel a bit nicer. Particularly the C1 and C2 buttons which have more travel. The rear dial feels more clicky as well. Which is a good thing.

Rear screen is on another level compared to the A1. I really like it. EVF is about the same. Although the 120fps refresh rate without resolution drop during focusing looks very nice. The A9III’s EVF might have a bit more aliasing though. Might be my imagination. Or maybe I have a setting wrong.

New C5 button is interesting. Feels weird to hold it down while using the camera when you want max burst. Maybe I just need to get used to it. For now, I went ahead and set it to burst toggle instead of burst hold.

Lots of new menu options I need to read up on. As pre usual, Sony naming choices are very cryptic 😂

Anyways, that’s all I got so far. Hope to put it through it’s paces soon

Edit: I had the EVF quality set to standard. Changed it to high quality and the aliasing is gone. Dumb mistake


Edited on Feb 10, 2024 at 11:14 AM · View previous versions
Feb 10, 2024 at 09:57 AM
randomguy
Offline
Details
p.2 #19A9III Thoughts from the first few hours
robert614 wrote:
Grip could still be taller IMO. Another 5-10mm would be perfect for me. With the new vertical grip attached though, the camera feels very nice in the hand.


Is the grip available in the US now? Here the grip is nowhere to be seen.
Feb 10, 2024 at 10:05 AM
robert614
Offline
Details
p.2 #20A9III Thoughts from the first few hours
randomguy wrote:
robert614 wroteGrip could still be taller IMO. Another 5-10mm would be perfect for me. With the new vertical grip attached though, the camera feels very nice in the hand.
Is the grip available in the US now? Here the grip is nowhere to be seen.


I preordered mine along with the A9III from B&H Photo.
They appear to still have them in stock. Not sure if they ship to Norway though

Feb 10, 2024 at 10:29 AM
FM ForumsSony ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister