Richard Marks
Rexel
newyorkone said:
I take your point, but DC Sang also makes a good point. Leica do not fully state exactly what they have altered and what the effects are. For example, improving one aspect may make something else slower or actually sacrifice another function.
My own thoughts are that the best skin tones came from the first update. (At least in relation to my software / printer / personnal tastes).This was done before the version that corrected for IR filters. Once a new download appears, leica remove the previous version. I actually have saved previous versions for the last 2 just in case what the new one does makes things worse. That way i could go back if i wanted to .
best wishes
Richard
palker
Established
"I actually have saved previous versions for the last 2 just in case what the new one does makes things worse. That way i could go back if i wanted to ."
Genius
Genius
cme4brain
Established
Red eye reduction excellent
Red eye reduction excellent
Here is an example of the new firmware's reduction of red eye with the M8 and Leica SF24D flash, done via shortened flash to shutter time. M8, crummy Peleng 8mm fish eye lens, self-portrait taken at arms length. No red eye in a picture where it should be, as can be seen in the 100% center crop of my eyes prone to red eye. Amazing.
Red eye reduction excellent
summilux said:
Here is an example of the new firmware's reduction of red eye with the M8 and Leica SF24D flash, done via shortened flash to shutter time. M8, crummy Peleng 8mm fish eye lens, self-portrait taken at arms length. No red eye in a picture where it should be, as can be seen in the 100% center crop of my eyes prone to red eye. Amazing.
Attachments
Richard Marks
Rexel
Red eye was not a problem with the SF24d at close range with the old software. At close range the the flash bulb is high enough above the eyes not to elicit the red reflex. You need to be about 10 feet away to get red eye. Its nothing to do with flash to shutter time. Any flash pointing directly at the pupil will do this! Light reflected back from the retina travels faster than any shutter!
I reckon you just got lucky.
I reckon you just got lucky.
cme4brain
Established
Richard Marks said:
Nope, I disagree and have examples. The delay in the interval between the flash and the shutter allowed the pupillary response to open again. While the flash tube is as high as it can get for the littel flash, I did get red eye on occasion at this distance. I did however get it more frequent at farther distances. I get it not at all now.
Richard Marks
Rexel
I am a bit sceptical
Red eye is a bit unpredicateble anyway
Anyhow if your happy thats great. I will certainly see if it improves with mine
incidentally the fisheye looks excellent
Best wishes
Richard
Red eye is a bit unpredicateble anyway
Anyhow if your happy thats great. I will certainly see if it improves with mine
incidentally the fisheye looks excellent
Best wishes
Richard
pstevenin
Established
After some testing, NR seems much better now form 640 up to 2500 which becomes more than usable, even in color.
WB is as crappy as usual. The camera froze for the first time (more or less 2000 actuations) . I just pull out & in the battery and everything is great now. Perhaps this is normal and nothing to do with the update.
It sounds worthy to implement it.
WB is as crappy as usual. The camera froze for the first time (more or less 2000 actuations) . I just pull out & in the battery and everything is great now. Perhaps this is normal and nothing to do with the update.
It sounds worthy to implement it.
Richard Marks
Rexel
Two pics with SF24 TTL mode (1) vs Auto mode (2)
The pupil is a bit smaller with the pre TTL flash, but there is no red eye on either. However if that is the total reduction in pupillary size from Pre flash I doubt it would be enough to significantly reduce red eye.
Time will tell!
Best wishes
Richard
The pupil is a bit smaller with the pre TTL flash, but there is no red eye on either. However if that is the total reduction in pupillary size from Pre flash I doubt it would be enough to significantly reduce red eye.
Time will tell!
Best wishes
Richard
Gabriel M.A.
My Red Dot Glows For You
I took some shots last night at ISO 1250, and I was so pleasantly surprised at how good underexposed shadows were, compared to how it was before. I'll take it for a spin at ISO 2500 and see how it plays out.pstevenin said:
ISO 1250 was very usable, but it looks like now it's as usable as ISO 640 previous to this firmware update.
Ben Z
Veteran
I use the SF24D but not in TTL so that part doesn't excite me, I almost never use the LCD...never for review and rarely to change a setting, so the scrolling isn't a big irritation (plus I've read many reports of people still getting the scrolling), the only Summarit that I might possibly maybe in a slim chance ever get is the 75mm and I won't care if it's recognized by the code, and I've found the battery level indicator as accurate as I need. So I'm probably going to pass or at least wait on this upgrade, because my M8 is working well with 1.107 and there's that old saying about not fixing what ain't broke 
pstevenin
Established
Ben Z said:I use the SF24D but not in TTL so that part doesn't excite me, I almost never use the LCD...never for review and rarely to change a setting, so the scrolling isn't a big irritation (plus I've read many reports of people still getting the scrolling), the only Summarit that I might possibly maybe in a slim chance ever get is the 75mm and I won't care if it's recognized by the code, and I've found the battery level indicator as accurate as I need. So I'm probably going to pass or at least wait on this upgrade, because my M8 is working well with 1.107 and there's that old saying about not fixing what ain't broke
Hi Benz,
If you shoot 1250 & 2500, it worth the trip just for the noise aspect and the fact you gain a stop as depicted by Gabriel MA.
Cheers Philippe.
Ben Z
Veteran
Philippe, the ISO control on a digital is like a radio volume knob, the higher you crank it the more noise you get. After some good advice and some experimentation I now always shoot everything @ ISO160 and "push process" it in ACR. I do all my DNG conversion in CS2 (gave up on C1 completely because I like the color much better in CS2 and I can turn the sharpening all the way off, and I like the interface and the fact I only need one program to do everything). The end result is far less noise than shooting @ higher ISO's and depending on the camera's NR. Any residual noise can be blitzed with Noise Ninja, but typically I don't have to until about ISO 1600. A big advantage of doing it this way is that I'm always using the lowest-possible "effective" ISO for the light, and I effectively get stepless ISO's, not just full-stop increments. It's very little additional work in post-processing considering the advantages.
Richard Marks
Rexel
Ben Z said:
I agree with BenZ here. 'Push processing' does seem ot have considerably less noise than pushing the ISO. The only probe with capture 1 is you can not push it a full 3 stops. Is this possible with your preferred software?
Richard
pstevenin
Established
Ben Z said:
Hi Benz,
I get the point but realised that internal NR is now acceptable at iso 1250 and even 2500 when correctly exposed without any postprocessing.
From 160 up to 2500 is 4 stops which I cannot correct in C1 or picture windows pro (but I will do some testing) nor compensate in A mode inside the camera ( Aperture priority may be some times useful). But I will do the test as suggested. If you have any procedure in mind, could you please share it?
It is just a matter of results acceptance and when I really need top quality, it usually means iso 160, tripod+cable release+lens stopped down a bit. For handheld shoots, I just think the results are just better straight out of camera (meaning DNG here) with this new release.
Philippe
PS: I do not own Photoshop (wayy to expensive for my straightforward photographic usage) did some tests with Lightroom, but prefer C1 + picture windows pro.
Ben Z
Veteran
Richard Marks said:
I use CS2 with the last Adobe Camera Raw plugin, version 3.7 so this applies only to that: the Exposure slider runs from -4 to +4 and according to the "Help" files, those equate to f-stops. There is an "auto" setting or you can set the default to 0 and DIY.
pstevenin said:
I hadn't thought about that. I admit to having not yet used A mode on my M8. I've used it on my Canon 20D but there I can compensate with the thumb wheel, whereas on the M8 I have to go into a menu. Even the M7 has a more user-friendly AE compensation.
Expose as if the ISO where whatever you want, adjust exposure to taste in ACR, remove any residual noise if necessary in Noise Ninja. My NN is the plugin-type so it runs inside CS2. Sometimes I have to play with shadows/highlights/curves etc but frankly I'm still boggled by Photoshop and lose interest quickly. If I spend more than 5 minutes post-processing a shot I get very bored
Elements 4.0 and 5.0 both accept the Camera Raw 3.7 plugin (which recognizes the M8), and 5.0 and 6.0 accept ACR 4.2, the same one used with CS3. So if you have Elements 4.0 or later, or an earlier Elements you can upgrade very inexpensively, you don't need Photoshop. Elements has been bundled with lots of cameras so chances are someone you know has a recent version on disc you could have. And Noise Ninja isn't very expensive. The stand-alone version is cheaper than the plug-in but not as convenient.
Last edited:
jaapv
RFF Sponsoring Member.
Ben Z said:
Depends on the subject matter, Ben. This works indeed very well on low to medium contrast photographs, but as you are effectively throwing away dynamic range in the process, it will result in not very good high-contrast shots, or at least not as good as they could be.
Gabriel M.A.
My Red Dot Glows For You
I have a hard time believing what Ben stated. Unless somebody is working with a firmware version where the high ISO is so bad, that doing this "push-processing" is not as bad, with that frame of reference.
I've seen underexposed shadows in ISO 160 where no matter how much "pushing" you do, they won't be as good as if you had exposed normally in a higher ISO (i.e. 2500)
Again, all of this may have to do with workflow, and that gives personal references of quality comparisons. I would never shoot at ISO 160 a shot which requires 1/8 sec. in ISO 2500. Again, different workflows, different ways of shooting. I've never ever ever, for example, used a tripod with my Leicas, but I'm sure there are some that hardly ever shoot without a tripod.
So what I'm saying is: I don't take photos, in the digital world, in such a way where I can see this 4-stop "push process"ing working where a "normally processed" image would be the same. But I'll try it; let's see if this alchemy works.
I've seen underexposed shadows in ISO 160 where no matter how much "pushing" you do, they won't be as good as if you had exposed normally in a higher ISO (i.e. 2500)
Again, all of this may have to do with workflow, and that gives personal references of quality comparisons. I would never shoot at ISO 160 a shot which requires 1/8 sec. in ISO 2500. Again, different workflows, different ways of shooting. I've never ever ever, for example, used a tripod with my Leicas, but I'm sure there are some that hardly ever shoot without a tripod.
So what I'm saying is: I don't take photos, in the digital world, in such a way where I can see this 4-stop "push process"ing working where a "normally processed" image would be the same. But I'll try it; let's see if this alchemy works.
Richard Marks
Rexel
Gabriel M.A. said:
Hi I get marginally less noise under exposing 3 stops (setting exposure compensation) -3 and then adding exposure in capture 1 than by pushing the ISO the same amount. Do try and see what you think. The suprising thing is that the playback still shows a reasonable image.
I take JAAPS point about potential problems with high contrast, but I consider a low light, high contrast situation something of a compromise any way.
Incidentally I do think the higher ISOs have improved with this update and there is less of a case for the BenZ route. I have specifically been shooting at 1250 today and this is remarkably good. Im just puzzled if Leica have made a substantial improvement in this why not at least mention it in their e-mail notice? Copuld it be that by so doing they admit to some 'flaw' in the previous version? Its beyond me.
Best wishes
Richard
Gabriel M.A.
My Red Dot Glows For You
Hmm, alright; I'll definitively try it.
OK, so this is "sepia", but I don't think I have any samples at my disposal this very moment. Shot at ISO 2500, hardly any noise (I did not apply any sort of color or lumina noise reduction); notice how the gradient to underexposed shadows starts showing the "grain"/noise. Certainly not bad. This was with firmware 1.092
The following is a crop (and in color!); no noise reduction either:
OK, so this is "sepia", but I don't think I have any samples at my disposal this very moment. Shot at ISO 2500, hardly any noise (I did not apply any sort of color or lumina noise reduction); notice how the gradient to underexposed shadows starts showing the "grain"/noise. Certainly not bad. This was with firmware 1.092
The following is a crop (and in color!); no noise reduction either:
Ben Z
Veteran
jaapv said:
What you say is true. However, blocked shadows that are smooth and almost noise-free may be more pleasing than more detailed but noisy shadows...or, they may not. Having shot slides for years and years, I'm comfortable with biasing exposure for highlights and letting the shadows go black if need be. In color I don't care for digital noise even if it's "film-like" because I didn't like grain in color film either, and if I have to choose between supressing noise or compressing dynamic range I'll choose in favor of less noise. Others may do the opposite, it's all good.
BTW Jaap, did you upgrade your firmware to 1.11 yet, and if so do you see an improvement in noise?