Noise

[Thommo]

Hey all,

Those of you who remember any of my postings on this forum may remember my photography is primarily live music. I have come to the point where I have outgrown my d70 and am looking to upgrade. The main issue with me upgrading is the need for less noise in my photos so here are a few questions.

1) how does the d200 compare to the d70 when it comes to low light noise. obviously it is going to be better but how much better.

2) does good quality glass effect the level of noise. I think i have been noticing this since beginning to use the 28-70 2.8 about 2 months ago, maybe i have just been shooting differantly.

3) why does canon have to have the one thing i want. low noise cameras, dont get me wrong i dislike everything else about canon cameras.

[Viz]

Nice to see that you are unashamedly biased.

I am in the same boat with a D70 and want something that will handle low light and longer exposures. I am very keen to hear what others say about noise performances. I honestly bought the D70 because of the feel of the body - have often thought about defecting to the canon camp for next body, but now I've comitted to the lens mount.

How about uploading a 30 second exposure of black at 100% crop at a fixed ISO for various cameras? Has this already been done somewhere?

[Yi-P]

How about uploading a 30 second exposure of black at 100% crop at a fixed ISO for various cameras? Has this already been done somewhere?

Great idea, will do that if I get to play around with other cameras.


The noise issue in the D70 is mainly in the dark/grey areas. I find shooting even at ISO800 is very acceptable noise if you overexpose the scene by about 1/2 to 2/3 stop (going 1 full stop is possible) and then lower the exposure in PP. Works good to remove noise at first shots.

[Viz]

Yes, overexposing is a work around, but you still have highlights, DOF and motion blur to contend with... which leaves me wanting a better sensor. I am personally quite scared because I have two people representing 2 parties at an event in 2 weeks that are both coming to me for photos... and it is all night shots with my D70s noise generator. I have access to an old manual fast wide F2 20mm (I think - have to check) and a 105VR (despite your great review turning others off - Yi-P), and the ol' kit... Noise levels has me scared.

And sorry Thommo - this is not answering your Q - hopefully it will aid ethereal, holistic conclusions

[PiroStitch]

Will the photos mainly be in b&w or colour? Usually if it's going to be b&w, bumping it up to ISO1250 is fine. Yes the D200 from what I've seen and heard handles noise better than the D70.

Unfortunately, Nikon still fall short in the noise department compared with Canon. Yes you can try to overexpose but like you said that's at the sacrifice of DOF or sharpness. I guess it really depends on what you're going for, more creative styling which includes movement or sharper pictures (ie. bugger all movement).

You could go with both camps and have both Nikon and Canon in your kit

[Viz]

I have to get band, equip, lighting, projections, pyros... so I need colour... in fact it is all about colour.

[xorl]

1) how does the d200 compare to the d70 when it comes to low light noise. obviously it is going to be better but how much better.

I haven't specifically compared my D200/D70 bodies quantitively, but my gut feel is the D200 gives a better print. The D200 has a higher resolution so the noise "averages out" a bit more, plus the built in NR helps.

Shooting low light gigs was one of the reasons I upgraded to the D200. Quite often they like to use heavy red lighting which throws the original D70 metering a bit. The RGB histograms and overexposure highlighting helps a lot to nail the exposures in these environments. AF and metering are significantly better on the D200 as well.

2) does good quality glass effect the level of noise. I think i have been noticing this since beginning to use the 28-70 2.8 about 2 months ago, maybe i have just been shooting differantly.

I generally use my 50/1.4 at gigs. The larger aperture is extremely helpful. f/2.8 usually seems like a bit of a stretch. I mostly end up around f/1.4 - f/2. I'd also consider getting a 35/2 or 85/1.8 depending on your style and the gigs that you attend.

3) why does canon have to have the one thing i want. low noise cameras, dont get me wrong i dislike everything else about canon cameras.

Have you tried 3rd party software like Noise Ninja? I've never felt the need myself, but some people find noise reduction software useful.

Also, I'd only ever judge the noise once the image is in it's final form instead of 100% on a monitor. If you are always printing 6x4 then a lot of what you see at 100% is irrelevant.

[Yi-P]

My rule of noise...

If it is noisy, turn it B&W... it looks great no matter how much noise you get...

[moz]

Seriously, the only real solution is to buy a decent camera. Since a Canon is apparently out of the question and the larger formats all top out at 800 ISO or less, you've shot yourself quite firmly in the head.

I use Noise Ninja with a 30D and it does tame the noise quite a lot. But since I mostly shoot for the web that's not really an issue - I can downsize and get quite a good one or two megapixel image pretty much regardless of the noise in the original.

I can get reasonable 4MP images at ISO6400 (from an 8MP/3200 nominal body) but they're noticably lacking compared to the 1DIIn that I've had the chance to look at. Whatever you do with the noise programs you do lose detail, and I've found that it can be hard to tell the difference between denoise-downsize-uprez and just denoising, the appearance of the detailed areas sometime looks better after the resizing. So test the noise progs quite hard before you buy, and remember that using them adds noticably to your workflow times.

[the foto fanatic]

Some people have reported that the D80 shows less noise than the D200. I have't used either, so I can't comment, but if I were in your position I'd check that out.

Good luck.

[Viz]

I just tried shooting black for 30 secs at 800 and 1600 ISO, but all you really get is black (some noticeable in the 1600 when zoomed in close) - I think we'd need controlled low light to see a reasonable benchmark of noise performance.

[moz]

I did a bunch of noise tests comparing the 300D vs 30D when I had both, it was one of the things that persuaded me to dump the 300D. In my opinion, comparing how well cameras perform with the lens cap on is missing the point - if you want this sort of test, I suggest fairly low light 9so you can get reasonable shutter/aperture combinations, then compare real shots. For bands, i'd be tempted to create a high contrast scene, even a high cntrast red scene if you can - get a lamp or two and dim out the room, then see what the shots look like.

Comparisons like this don't have to be incredibly precise if you're only looking at one thing, it's only when a million measurebators are having underwear incidents over it that you have to be perfect. If you wanted a "real" comparison, take both cameras and the relevant lenses to a gigi and use the owner(s) of the borrowed gear as assistants. They hold the toys, change the lenses, and advise on settings. You take the shots. That way no-one should lose their gear, and you get a real-life feel for how stuff performs.

I suspect that you could even arrange an outing to some cheap'n'cheerful band if you wanted a drag out some of the people here that have nice toys.

Say, Sunday at the forensic gig? You don't have to especially like the music, just go along and take photos to see how it works.