Why L 0.3, 0.7 1.0? - FM Forums

Home · Register · Search · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda

Pre-order new Rokinon 16mm f/2.0 ED AS UMC lens

Pre-order new Sigma 18-35mm f/1.8 DC HSM lens

Bower 35mm f/1.4 price dropped to $379

FM Forums \| Nikon-mount SLRs \| Join Upload & Sell
1 2 end

Why L 0.3, 0.7 1.0?

480sparky
Offline
• •
Upload & Sell: Off

p.1 #1 · Why L 0.3, 0.7 1.0?

My response to "Why do Nikons use H 0.3, 0.7, 1.0 and 2.0 instead of ISO 12,500, 16,000 etc." were always answered with, "Because the ISO (International Standards Organization) does not recognize any film or digital speed higher that 10,000. So to avoid getting into trouble, Nikon uses the H designations."

Well, my theory worked until I got my D600. And instead of finding ISOs of 50, 64 and 80 (familiar ASA numbers from my film days), I find L 0.3, 0.6 and 1.0.

Obviously, ISO speeds of 50, 64 and 80 exist and are perfectly acceptable by the ISO. So why use the L system instead of actual ISO ratings that have existed for years?

Jan 03, 2013 at 03:28 PM

Danner
Offline
• •
Upload & Sell: Off

p.1 #2 · Why L 0.3, 0.7 1.0?

Because those digital gain settings are operating outside of the designed ISO ratings of the sensor system, and result in poorer image performance. Use with caution.

Jan 03, 2013 at 03:32 PM

NathanHamler
Offline
• • •
Upload & Sell: Off

p.1 #3 · Why L 0.3, 0.7 1.0?

B/c the sensor/camera doesn't NATIVELY go past 6400, or lower than 100.....to get past that the camera is basically pushing or pulling the exposure just like you would in Photoshop, or Lightroom, or Aperture.....i guarantee it has nothing to do with the ISO not recognizing past iso 10,000.....the D3s Natively goes to 12,800, so if you dial it to 12,800, it will say 12,800.

Jan 03, 2013 at 03:33 PM

480sparky
Offline
• •
Upload & Sell: Off

p.1 #4 · Why L 0.3, 0.7 1.0?

Yet I still have ISOs of 1600, 3200, 6400, and 8000. Why not start the H designations there?

Jan 03, 2013 at 03:35 PM

BenV
Offline
• • • • •
Upload & Sell: On

p.1 #5 · Why L 0.3, 0.7 1.0?

480sparky wrote:
Yet I still have ISOs of 1600, 3200, 6400, and 8000. Why not start the H designations there?

I don't think you understood what some of the people are saying. It's pushing the sensor past Nikons standards, and Nikon doesn't 'support' that, but they give the option to do so if needed. Any of the H iso's are not recommended to use by Nikon.

Jan 03, 2013 at 04:42 PM

Nathan Padgett
Offline
• •
Upload & Sell: On

p.1 #6 · Why L 0.3, 0.7 1.0?

NathanHamler wrote:
B/c the sensor/camera doesn't NATIVELY go past 6400, or lower than 100.....to get past that the camera is basically pushing or pulling the exposure just like you would in Photoshop, or Lightroom, or Aperture.....i guarantee it has nothing to do with the ISO not recognizing past iso 10,000.....the D3s Natively goes to 12,800, so if you dial it to 12,800, it will say 12,800.

Sometimes I think underexposing just a little in the higher ISO's and bumping it up later in post gives better results than a shot with the higher ISO to begin with (especially the "H" settings). Anyone else feel the same?

Jan 03, 2013 at 05:32 PM

Nathan Padgett
Offline
• •
Upload & Sell: On

p.1 #7 · Why L 0.3, 0.7 1.0?

Back on topic.....in the future if ISO continue to climb. I wouldn't mind seeing some type of new system for labeling the sensor sensitivity. The math is gonna get too crazy with 12,800 - 25,600, 51,200 - 102,400 - heck another 4 stops and we are over 1,000,000!

Lets call it "DSS" (Digital Sensor Sensitivity) for now. Instead of starting at 100, we start at 1. I guess it would still have to double each stop, or would simply going 1-2-3-4 work without confusing too many photographers used to the ISO standards? You could have 1/3 stops as 1.3 - 1.6 ect.

Jan 03, 2013 at 05:42 PM

480sparky
Offline
• •
Upload & Sell: Off

p.1 #8 · Why L 0.3, 0.7 1.0?

Nathan Padgett wrote:
.........
Lets call it "DSS" (Digital Sensor Sensitivity) for now. Instead of starting at 100, we start at 1. I guess it would still have to double each stop, or would simply going 1-2-3-4 work without confusing too many photographers used to the ISO standards? You could have 1/3 stops as 1.3 - 1.6 ect.

ISO started out at like that, back when it was ASA. But they didn't use a linear system.

We can easily revert to the DIN or Log system, which is precisely what you've described ('cept each number is 1/3 stop, so adding 3 would increase the speed by one stop).

Jan 03, 2013 at 05:52 PM

NathanHamler
Offline
• • •
Upload & Sell: Off

p.1 #9 · Why L 0.3, 0.7 1.0?

The "native" iso range is whatever is published in the specs, so a d600 is 100-6400, and a d3s is 200-12,800

Jan 03, 2013 at 05:53 PM

NathanHamler
Offline
• • •
Upload & Sell: Off

p.1 #10 · Why L 0.3, 0.7 1.0?

Nathan Padgett wrote:
NathanHamler wrote:
B/c the sensor/camera doesn't NATIVELY go past 6400, or lower than 100.....to get past that the camera is basically pushing or pulling the exposure just like you would in Photoshop, or Lightroom, or Aperture.....i guarantee it has nothing to do with the ISO not recognizing past iso 10,000.....the D3s Natively goes to 12,800, so if you dial it to 12,800, it will say 12,800.

Sometimes I think underexposing just a little in the higher ISO's and bumping it up later in post gives better results than a shot with the higher ISO to begin with (especially the "H" settings). Anyone else feel the same?

Actually i find the opposite...i'd rather overexpose and pull it back for less noise...

Jan 03, 2013 at 05:57 PM

JimFox
Online
• • • • • • •
Upload & Sell: On

p.1 #11 · Why L 0.3, 0.7 1.0?

Nathan Padgett wrote:
NathanHamler wrote:
B/c the sensor/camera doesn't NATIVELY go past 6400, or lower than 100.....to get past that the camera is basically pushing or pulling the exposure just like you would in Photoshop, or Lightroom, or Aperture.....i guarantee it has nothing to do with the ISO not recognizing past iso 10,000.....the D3s Natively goes to 12,800, so if you dial it to 12,800, it will say 12,800.

Sometimes I think underexposing just a little in the higher ISO's and bumping it up later in post gives better results than a shot with the higher ISO to begin with (especially the "H" settings). Anyone else feel the same?

Hey Nathan,

My experience has been the exact opposite. My cleanest files are always when properly exposing at any given ISO. So to underexpose at ISO 3200 and then overexpose when processing will look worse then simply exposing properly at ISO 4000 or ISO 5000, etc. It sounds like your experience has been different. But for me, I shoot a lot of low light shots, so I did my own tests for myself, and that was my conclusion.

Jim

Jan 03, 2013 at 11:58 PM

Rodolfo Paiz
Offline
• • • • •
Upload & Sell: On

p.1 #12 · Why L 0.3, 0.7 1.0?

Nathan Padgett wrote:
Sometimes I think underexposing just a little in the higher ISO's and bumping it up later in post gives better results than a shot with the higher ISO to begin with (especially the "H" settings). Anyone else feel the same?

No.

Adding some detail to that, read up on ETTR (Expose To The Right) and why it works. Underexposing in camera and then pushing in post will always give you more noise and less image quality. When using extremely high ISO, you need to either nail the exposure in camera or slightly overexpose in camera, then adjust down in post. Overexpose then pull, never underexpose then push.

Jan 04, 2013 at 12:05 AM

EB-1
Online
• • • • • • •
Upload & Sell: Off

p.1 #13 · Why L 0.3, 0.7 1.0?

480sparky wrote:
My response to "Why do Nikons use H 0.3, 0.7, 1.0 and 2.0 instead of ISO 12,500, 16,000 etc." were always answered with, "Because the ISO (International Standards Organization) does not recognize any film or digital speed higher that 10,000. So to avoid getting into trouble, Nikon uses the H designations."

Well, my theory worked until I got my D600. And instead of finding ISOs of 50, 64 and 80 (familiar ASA numbers from my film days), I find L 0.3, 0.6 and 1.0.

Obviously, ISO speeds of 50, 64 and 80 exist and are perfectly acceptable by the ISO. So why use the L system instead of actual ISO ratings that have existed for years?

The explanation is ridiculous. If enough people complain perhaps they will change it in the next model.

EBH

Jan 04, 2013 at 12:15 AM

Rodolfo Paiz
Offline
• • • • •
Upload & Sell: On

p.1 #14 · Why L 0.3, 0.7 1.0?

480sparky wrote:
My response to "Why do Nikons use H 0.3, 0.7, 1.0 and 2.0 instead of ISO 12,500, 16,000 etc." were always answered with, "Because the ISO (International Standards Organization) does not recognize any film or digital speed higher that 10,000. So to avoid getting into trouble, Nikon uses the H designations."

Whoever told you that simply made something up that sounded logical instead of admitting that they didn't know the real reason. The reality is that any given sensor can only reach a certain range of ISO settings. Nikon refers to those ISO settings with numbers. However, sometimes you need either a lower or higher ISO than what the sensor can really reach, so Nikon offers the "L" and "H" ranges.

A D800, for example, has a native ISO range of 100 to 6400. The sensor cannot do any better than that. So, to shoot at "H1.0" (ISO 12800) the camera takes a shot at ISO 6400, but with a shutter speed that's twice as fast as it would have otherwise used. The resulting image is underexposed by one stop, and then the camera pushes the exposure by one stop to compensate.

That's why the camera refers to "H1.0" instead of ISO 12800: because you're pushing the best that the sensor can give you by one stop. And that also explains why both the "L" and "H" ranges offer reduced dynamic range compared to their nearest numbered ISO value. But you may still want to use them sometimes... knowing how they work, and what their pros/cons are, is part of knowing your tools as best you can to get maximum benefit from them.

Jan 04, 2013 at 12:15 AM

Rodolfo Paiz
Offline
• • • • •
Upload & Sell: On

p.1 #15 · Why L 0.3, 0.7 1.0?

EB-1 wrote:
The explanation is ridiculous. If enough people complain perhaps they will change it in the next model.

They won't, nor should they. Knowing where the sensor's native capability ends, and software pushing/pulling begins, is critical information for the photographer to make the right choices. The explanation given to the OP is ridiculous, but it's also false. The real reason is both technical and valid.

Jan 04, 2013 at 12:17 AM

NathanHamler
Offline
• • •
Upload & Sell: Off

p.1 #16 · Why L 0.3, 0.7 1.0?

JimFox wrote:
Nathan Padgett wrote:
NathanHamler wrote:
B/c the sensor/camera doesn't NATIVELY go past 6400, or lower than 100.....to get past that the camera is basically pushing or pulling the exposure just like you would in Photoshop, or Lightroom, or Aperture.....i guarantee it has nothing to do with the ISO not recognizing past iso 10,000.....the D3s Natively goes to 12,800, so if you dial it to 12,800, it will say 12,800.

Sometimes I think underexposing just a little in the higher ISO's and bumping it up later in post gives better results than a shot with the higher ISO to begin with (especially the "H" settings). Anyone else feel the same?

Hey Nathan,

My experience has been the exact opposite. My cleanest files are always when properly exposing at any given ISO. So to underexpose at ISO 3200 and then overexpose when processing will look worse then simply exposing properly at ISO 4000 or ISO 5000, etc. It sounds like your experience has been different. But for me, I shoot a lot of low light shots, so I did my own tests for myself, and that was my conclusion.

Jim

Just to be clear, you're responding to Nathan PADGET, and not Nathan HAMLER, correct? lol..... Obviously it's always best to nail the exposure, but as Rodolfo suggested, ETTR (exposing to the right), or slightly over exposing, and then pulling it back in post, always yields cleaner results in terms of noise at higher iso's, but maybe not in terms of dynamic range and preserving highlights.....although, most modern sensors still hold highlights fairly well even at higher ISO's...

Jan 04, 2013 at 12:29 AM

480sparky
Offline
• •
Upload & Sell: Off

p.1 #17 · Why L 0.3, 0.7 1.0?

Rodolfo Paiz wrote:Whoever told you that simply made something up that sounded logical instead of admitting that they didn't know the real reason.............

No one told me this. I figured it all by my lonesome. The ISO does not recognize ISO over 10,000.

Jan 04, 2013 at 12:35 AM

NathanHamler
Offline
• • •
Upload & Sell: Off

p.1 #18 · Why L 0.3, 0.7 1.0?

480sparky wrote:
Rodolfo Paiz wrote:Whoever told you that simply made something up that sounded logical instead of admitting that they didn't know the real reason.............

No one told me this. I figured it all by my lonesome. The ISO does not recognize ISO over 10,000.

Just cause they dont recognize it, doesnt mean it doesnt exist...it's just a number....doesnt change the physics of sensors...

Jan 04, 2013 at 02:31 AM

Rodolfo Paiz
Offline
• • • • •
Upload & Sell: On

p.1 #19 · Why L 0.3, 0.7 1.0?

480sparky wrote:
No one told me this. I figured it all by my lonesome. The ISO does not recognize ISO over 10,000.

Your original post read quite differently to me. Nikon does not seem to be worried about the International Standards Organization in any case, since their reasons for using H and L ranges are as explained above: those are the ISO values that are achieved by pushing or pulling the exposure beyond what the sensor is natively capable of producing without manipulation.

Jan 04, 2013 at 04:07 AM

mfletch
Offline
• • •
Upload & Sell: On

p.1 #20 · Why L 0.3, 0.7 1.0?

480sparky wrote:
Rodolfo Paiz wrote:Whoever told you that simply made something up that sounded logical instead of admitting that they didn't know the real reason.............

No one told me this. I figured it all by my lonesome. The ISO does not recognize ISO over 10,000.

Regardless of where you got this info, and whether it's true or not, it has nothing to do with the use of L and H ISOs.

The reason for these designations have been clearly explained in this thread.

Jan 04, 2013 at 04:56 AM

1 2 end
FM Forums \| Nikon-mount SLRs \| Join Upload & Sell 1 2 end