Jemini wrote:
And the verdict? Sorry, don't understand German.
Nothing extraordinary. He compares different noise reduction strengths in camera in raw, then later in jpg.
I don't like that he compares at 200% each, he should equalize output size like Dpreview is doing it. Based on his 200% comparisons his threshold is iso 6400, but I think that to be worthless unless parameters for comparison are equalised.
Looking at some of the RAWs in LR, the A1 at 25,600 is retaining finer detail than the A7RIV at 12,800. Although the German reviewers really messed up the exposures....I'm not sure why they didn't keep them consistent? So I'm having to make 1/3 or 2/3 adjustments to exposure in LR in order to equalize the settings and that may be throwing off the results as it is the A7RIV files that are underexposed compared to the A1.
baltmin wrote:
I opened them in raw therapee and I can see the A1 being 1 stop better than the A7RIV and close to a9.
I will open them in imaging edge as well and I will update you, just in case raw therapee does some kind of magic (which I like!)
The german speaking guy in the clip says the same thing: one stop better than A7R4.
arbitrage wrote:
Looking at some of the RAWs in LR, the A1 at 25,600 is retaining finer detail than the A7RIV at 12,800. Although the German reviewers really messed up the exposures....I'm not sure why they didn't keep them consistent? So I'm having to make 1/3 or 2/3 adjustments to exposure in LR in order to equalize the settings and that may be throwing off the results as it is the A7RIV files that are underexposed compared to the A1.
From your comparisons I have to say the A1 really looks great, esp. compared to the A7riv. The A9 has an AA filter helping it further, but you can see the tiny letters below the shaded man in the iso3200 comparison being blurred strongly on the A9 shot. The plastic of the color chart is strongly smoothed, too, revealing clearly less detail compared to the A1.
The yellow spice in the iso8000 shot: If you zoom in in the top area of that bowl the A9 shot is clearly stronger blurred out than the A1 shot, so there is even more headroom for additional subtle noise reduction.
At all the isos relevant to _me_, the A1 shows a very strong performance. I hoped for that, but didn't expect that.
I'm wondering about the color science of the A1 vs a9ii looking at the sample photos. In the bread photo there seems to be less saturation in the orange/red in the a1 photo. Most of the others, too - yet the photo of the currency seems to have more saturation in the red/magenta. I don't know if that's just a camera setting/WB issue or if the colors are accurate. I've noticed it in some of the other sample photos from the a1 that have been shown in videos and articles, too. Although there may be no good profile for the camera yet, so I'm just trying to wrap my head around things.
arbitrage wrote:
Looking at some of the RAWs in LR, the A1 at 25,600 is retaining finer detail than the A7RIV at 12,800. Although the German reviewers really messed up the exposures....I'm not sure why they didn't keep them consistent? So I'm having to make 1/3 or 2/3 adjustments to exposure in LR in order to equalize the settings and that may be throwing off the results as it is the A7RIV files that are underexposed compared to the A1.
Yeah, I downloaded the color chart ISO series and went straight to ISO 100 in hopes of making a comparison of base ISO dynamic range through RawTherapee. I don't think it's possible to do effectively for a few reasons. First, the A1 seems to have been exposed 2/3 of a stop brighter assuming that their ISO 100s are equally sensitive (they might not be, of course). I then opened both RAWs in the editor and found that with RawTherapee defaults the relative saturation levels were way off: the A7RIV file was much more saturated. That made the A1 look superior in terms of shadow noise. I am not an expert with this software but I tried to match them visually, though it was imperfect. I added 3.5 stops of EC to each file and brought back the highlights as much as possible then exported to TIFF to compare.
With all that done, the A7RIV file seemed noticeably cleaner in the blacks of the chart, although they looked very similar in anything colorful like the color checker patches or the bills. That was a bit surprising given the 2/3 stop advantage in shutter speed with the A1, but again there are too many variables. For this reason I hesitate sharing the crops here.
If you have extra time on your hands, thoughts on opening the ISO 100 shots (3428 for the A1 and 8407 for the RIV), disabling NR, and boosting brightness to see how they compare? It would at least help counteract one of the variables in my examination.
ketang wrote:
First, the A1 seems to have been exposed 2/3 of a stop brighter assuming that their ISO 100s are equally sensitive (they might not be, of course).
Just FYI, it doesn't matter if any camera s ISOs are equally sensitive. Controlling the total light going into the camera (f-stop, or rather t-stop with the same lens and shutter speed) and then matching exposure in post has been the known standard forever. It's bizzare that a review site would muck that up.
If you watch Gordon Laing’s video comparing the a1 to R5 you’ll notice a magenta cast in the highlights of the a1’s image of the tulip. Compare that to the R5 image which has no magenta.
At least it appears so.
saxguy wrote:
I'm wondering about the color science of the A1 vs a9ii looking at the sample photos. In the bread photo there seems to be less saturation in the orange/red in the a1 photo. Most of the others, too - yet the photo of the currency seems to have more saturation in the red/magenta. I don't know if that's just a camera setting/WB issue or if the colors are accurate. I've noticed it in some of the other sample photos from the a1 that have been shown in videos and articles, too. Although there may be no good profile for the camera yet, so I'm just trying to wrap my head around things. ...Show more →
saxguy wrote:
I'm wondering about the color science of the A1 vs a9ii looking at the sample photos. In the bread photo there seems to be less saturation in the orange/red in the a1 photo. Most of the others, too - yet the photo of the currency seems to have more saturation in the red/magenta. I don't know if that's just a camera setting/WB issue or if the colors are accurate. I've noticed it in some of the other sample photos from the a1 that have been shown in videos and articles, too. Although there may be no good profile for the camera yet, so I'm just trying to wrap my head around things. ...Show more →
If you are talking about RAW files, be mindful that the color profile for the camera might not be fully done yet.
There might also be a change in Auto WB on Sony, if that was used. I often feel that Sony Auto WB tends to be a bit on the green side in the Tint slider, so I wouldn’t be surprised to that balance shifted a touch to the Magenta side. Easily adjusted in the RAW converter of course but still something to keep an eye on if you are coming from an older Sony camera without the new WB sensor on the A1 and A7S III.
