darrenmambo

<p>In my opinion, "Are they worth buying" is the same answer as "will you ever shoot anything over ASA 400?"

<p>I've been using Neat Image for years now... I recently downloaded the trial version of Noise Ninja, and tested them both on the same images.

<p>My general finding was that (Neat Image, Edit->Fade 80-90%) is pretty much identical to Noise Ninja.

<p>As to the quality, it really depends on how much time you want to spend profiling. Both programs let you make "custom" profiles for your particular camera, but what I've learnt is that the profile you make at home under normal "home" lighting (e.g. incadescent, flourescent) will not be perfect for the shots you take in a club lit entirely by coloured spotlights, or industrial floodlights, etc. Additionally, every change you make in Camera Raw effectively changes the noise profile for that particular image.

<p>So, long story short, what I've learnt over time is that if you're after top quality, it's best to (A) either treat every shoot like a catalogue photographer would, and shoot your GretagMacbeth colour chart and your noise target under the same lighting as everything else, and/or (B) be willing to create a custom profile from every image you shoot.

Are you looking to pose the dancers, or just take photos of people dancing?

If the former, talk to the professionals at a dance studio. See if they're interested in TFP (time-for-prints) where they volunteer their time in exchange for your prints.

If it's social, I'd very strongly recommend you find a tango studio and sign up for lessons. Good dance photography is very much about understanding the movements and the patterns, so that you can anticipate that if they are dancing pattern X, they will be in a good position for a photo on count Y.

Additionally, if you take some lessons, and go to the social dances for a while without your camera, people will get to know you, and will be much more relaxed about you taking photos of them.

Sorry, clarification on one of the "tips"

<p>Photograph from the right side of the man when the dancers are in open position, but from the left side when they're in closed position.

<blockquote><i>

When i asked about how i should approach organizers i was told the only way i could do it would be to sell to the dancers. When i mentioned this after the event i got a most viscious reply saying i was banned from future events.

</i></blockquote>

<p>Sounds to me like a simple case of "someone forgot, or chose to forget, what they said."

<p>Generally, if you're going in to try and sell photos:

<ol>

<li>Talk to the organizers a month or two ahead of the event.

<li>Get a contract. Or at least something written. Include your prices, and arrange a kickback to the organizers (e.g. 10% of your gross sales.) Have someone there sign it; preferably two people.

<li>Talk about pre-marketing; see if they'll add something to the program or signage saying "professional photographs will be available from Manuel's Photo" or something.

<li>Call the organizers a day or two ahead, remind them who you are, and ask them for "early access" to set up your gear, and hopefully watch a rehersal if possible. Find your lines, and leave your tripod in place, otherwise people will take your spot.

<li>When there, be professional. Of course, that doesn't need to be said.

</ol>

<p>Regardless of whether or not you feel dancing is a sport, you have to approach it as sports photography. In other words, if you don't know anything about "the sport," your photos will not be as good as if you can predict what's going to happen next. For example:

<ul>

<li>Most of the flair in cha-cha will be on beat 4, but on beat 3 in rhumba

<li>As a general rule, always photograph the right side of the man; that way, the lead arm won't be blocking your shot.

<li>Learn about musicality and phrasing; music is even more predictable than dancers, and good dancers dance to the music, not just the beat.

</ul>

<p>Semi-obligitory street credentials ;)

<a href="http://www.mambo.net/as/imgsearch/auto/ballroom+dance">

Ballroom</a>

and

<a href="http://www.mambo.net/as/imgsearch/auto/lindy">Lindy Hop

</a>

pictures

<p>Cheers,

<br>Darren

Thank you for all the responses. I print on a 2200, and had wondered if I could save some money with the "regular" matte paper. It sounds like no, at least for work that I care about.

Cheers,

-DArren

Does anyone know the difference between Epson's Matte paper compared

to their Enhanced Matte paper? Other than the fact that Enhanced Matte

is about 40% more expensive?

<p>Thanks to everyone for their help.

<p>I've put together a web page about how I'm working around this problem, including the printer test page that I used to determine my printer's black and white points with the profile.

<p><a href=http://www.mambo.net/as/view/115>http://www.mambo.net/as/view/115</a>

<p>Cheers,

<br>Darren

My question is on the interaction between calibration and targeting...

My example is using Photoshop CS.

<p>I have calibrated my printer, an Epson 2200, using Monaco EZColor.

However, if I print a black step wedge with source space=Adobe RGB and

target space=the printer profile I created, all the blacks between RGB

0 and 32 print as the same tone (i.e. Zone 0) - The first visible

graduation is between 32 and 34.

<p>To "rescue" my shadows, I then have to target the image for

printing. I do this by using a levels (or curves) adjustment layer,

and either (a) using the black eyedropper with it's value set to 32,

or (b) dragging the output level black slider to 32.

<p>White does not have this problem - there is a visible difference

between 253 and 255.

<p>Now my questions are:

<ol>

<li>Given that I'm using a profile, should I not need to target at

all? In other words, should the profile include the fact that the

printer can't reproduce differences in tone below RGB 32?

<li>By targeting using an adjustment layer, am I upsetting the profile

itself by compressing its dynamic range?

<li>Am I doing something wrong here? Or am I misunderstanding something?

</ol>

<p>Cheers,

<br>Darren

I'd also recommend "The Camera Store" on 11th Ave & 7th St. SW.

All the large photo stores here are in an eight block radius. The other "big" ones are:

Vistek: Around 10th Ave & 12th St. SW. Note the sign says "West Canada Graphics & Vistek." Calgary branch of Canada's largest semi-pro/pro photo chain. If they don't have something, they can get it from their Toronto store in 2-3 days.

Saneal Cameras: On 11th Ave right east of 14th St. SW. They also sell some VCR's and such stuff, which always made me wonder "who would go to a camera store to shop for a VCR?"

Many different ways to do this:

1.

Create a new layer above the image in mode "soft light," and fill with neutral 50% grey. Select an airbrush, about 5% flow and 10% opacity. Set brush colour to white, and gently paint out the stains.

2.

Lasoo the teeth. Use "selective colour" and select the yellows to get a mask. Make a curves adjustment layer and pull it down.

3.

Use the clone stamp tool to clone the good teeth onto a layer above the bad teeth. Decrease opacity to make it look natural.

NeatImage can give some really great results, but unfortunatly, like so many software programs, it presupposes that you have some understanding of the underlying algortithms.

After playing with it for some time, I've found settings that work well. But I still usually layer the cleaned image over the original and set transparancy to 50-75% to avoid that "artificial" look.

Funny how we have to re-introduce a tiny bit of the same noise that we were trying to remove in order to have it look natural.

People who don't think tape is a viable solution for backing up corporate networks have obviously never had to set up a backup system that can handle backing up a 40 Tb RAID network...

But for the home, I second the removable USB disk option as well.

One thing I'm doing to help get over the embaressment / nervousness is setting myself little photo projects. My most recent one was "People Carrying Things." I just sat myself on some stairs at a street corner, in full view, and took pictures of people carrying things when they walked by.

By doing this, I find I just kind of blend into the background of the cityscape. The eye notices movement, so I didn't really move much. I'd watch the distance for people coming who were carrying things, then pre-compose the shot where I expected them to pass. I then hold the camera in place and open both eyes, so I can see them coming. When they walk into frame, a quick re-compose if necessary, snap, and keep the camera up; don't drop it right away and it's not obvious they noticed you taking their picture.

Seems to work well. I find the more of these I do, the more comfortable I am actually wandering around and snapping pictures on the move.

I rented a film scanner (Polaroid SprintScan) and spent a whole weekend earlier this month just scanning b&w negatives as 'Generic Negative/16 bit'... Man, this takes forever; only got through about one year's worth. argh.

<p>However, I quickly learnt that it takes even longer to load them all in and tune them. Each channel usually has 'something' that I like and something that I don't. For example, in portraits, the red channel has such a clear rendering of skin tone but is very soft. The Blue channel has a harsh rendering of skin tone but usually good background detail. Green is usually somewhere in between, and it varies for every single image...

<p>I started by playing with channel mixer, but it's kind of annoying because it's hard to compare whether 60/30/10 looked better than 50/30/20, since you can't just flip between them...

<p>So I created an action to combine channels and let me choose one for further tweaking. My so-called workflow is then:

<p><ul><li>

Load image

<li>Manually drag endpoints in each 16 bit colour channel (R,G,B) so that I'm not clipping anything (manually, since auto-levels seems to do a crappy job.)

<li>Tweak the RGB pseudo-channel in levels since above moved it around. Set endpoints and move middle slider (gamma) if desired.

<li>Run Action, which contains:

<ol><li>Convert to 8-bit mode

<li>Calculations: Multiply Red by Green to get new channel

<li>Calculations: Multiply Red by Blue to get new channel

<li>Duplicate Red Channel

<li>Auto-Stretch contrast on Red Copy Channel (to give different values)

<li>Calculations: Multiply Red Copy by Green to get new channel

<li>Calculations: Multiply Red Copy by Blue to get new channel

<li>Calculations: Multiply Green by Blue to get channel GB

<li>Calculations: Multiply Red by GB to get new channel

<li>Calculations: Multiply Red Copy by GB to get new channel

<li>Delete channel Red Copy

<li>Delete channel GB

</ol></ol>

<p>This gives me six different "basic interpritations" of the image in greyscale that I can quickly compare just by clicking on them in the channels pallete. For most images, the middle two are the best, but not all. I then pick one, tweak it if necessary, then convert to greyscale to get rid of the others.

<p>Disclaimer: I'm still a beginner. If anyone has any better ideas, I'd be happy to hear them.

Very Cool.

In the interests of feaping creaturism, how about the ability to do all these by "category," e.g. only Portraits or Landscapes?

Brian wrote: <i> Did you know that our "image-display" URL is the third highest entry URL into our site?</i>

<p>Sounds like a fairly easy solution...

<pre>

IF (REFERER != www.photo.net)

DELETE PHOTO

</pre>

Of course, that'd be a nice easy route to abuse as well... Perhaps

<pre>

IF (REFERER == www.ebay.com)

REDIRECT TO (Jpeg saying "Would you buy a used car from this person?")

DELETE PHOTO

</pre>

<p>But seriously...

<p>1. If you decide to go towards per-user quotas (which I think is quite fair,) I think <b>history (i.e. time)</b> is probably the best predictor... Have you contributed to discussions (i.e. like this?) over time, not

just in one massive burst? Have you given critiques in the past, not

just now?

<p>Granted, this implies a "bias" towards existing, active users, a need to "pay your dues," but then again, so do most things in life.

<p>2. Change policy so that all critique requests have three-week life before being auto-deleted, unless enough people mark it as "worthy of saving." Of course, then we're back to the whole "Who do

you believe" problem...

<p>3. Simple one: No pay, photos no stay, and learn to deal with it.

I'm a somewhat low rater - I try to follow Phillip's guidelines and I generally only rate photos that "interest" me. My average rating (of others) is 5.78/5.9. The highest rating I have ever given was

<a href=http://www.photo.net/photo/513920>9/9</a>. To my way of thinking, giving a "10" is something you would do only once in your life.

<p>Since p.n already calculates "average rating given," how about normalizing the ratings based on that? If someone has an average of 8/8 and rates a photo as 7/7, that becomes 4/4 (normalizing around 8.)

Of course, you'd use a more mathmatically robust algorithm in real life.

<p>I've often thought there's a really good masters' thesis just waiting to happen on normalizing community-based ratings...

<p>Disclaimer: I have never uploaded a photo for critique because after about 500 rolls, I still see my photography in the 5/5 or 6/5 range and there's plenty of that already on photo.net