Instagram carousels

01 · The short version

TL;DR — what to actually do

02 · Limits and formats

Technical limits & formats (verified 2026)

Slide count

• Maximum: 20 photos/videos per carousel, increased from 10 on 8 August 2024.
  • The Verge: “You can now share 20 slides in one Instagram post” — theverge.com/2024/8/8/24216111
  • Corroborated by MacRumors (same day, cites The Verge) — macrumors.com/2024/08/08/instagram-20-photos-carousel-posts and Metricool (rollout notes; some accounts still saw 10 during gradual rollout) — metricool.com/instagram-increases-carousel-content-limit
  • Rollout began Aug 2024; by 2026, 20 slides is standard. Caveat: some third-party schedulers still cap carousels at 10 slides (Sprout Social's own guide says its app allows 10) — for a full 20-slide post, publish from the Instagram app. (sproutsocial.com/insights/instagram-carousel)
• Minimum is 2 slides. The count is combined across photos and videos (e.g., 18 photos + 2 videos).

Aspect ratios & the cover-crop rules

• Feed carousels support 1:1 (square), 4:5 (portrait), 1.91:1 (landscape), and — since May 2025 — 3:4 portrait (the default phone-camera ratio), announced by Adam Mosseri: “If you upload a 3:4 image, it'll now appear just exactly as you shot it” (single image or carousel). Source: 9to5Mac, 29 May 2025 — 9to5mac.com/2025/05/29/instagram-changes-standard-photo-aspect-ratio
• Meta's own design guidance for feed media: ratios from 1.91:1 up to 9:16 are accepted, but taller media gets auto-cropped to the recommended 4:5; upload at least 1080×1080 px; “no maximum resolution”. (Meta Business Help Center, “Design Requirements for Instagram Feed Ads”, facebook.com/business/help/430958953753149 — this is the ads spec page; it matches what all major tools report for organic feed too.)
• The first slide sets the crop for every other slide. If slide ratios differ, Instagram auto-crops to the first slide's ratio — and per Buffer, “if Instagram auto-crops your carousel images, you won't be able to adjust that crop later.” You can also choose a “mixed dimensions” mode, but it adds awkward spacing around non-matching slides. Practical rule: crop every slide to the same ratio yourself before uploading. (Buffer, Instagram image size guide 2026 — buffer.com/resources/instagram-image-size; same rule stated by Sprout, Hootsuite, Later.)
• Where the cover gets cropped: in the feed, the post shows at its own ratio (up to 4:5). In the profile grid, previews now display at 3:4 (Instagram moved from the square grid in early 2025), so a 4:5 cover loses a strip on the left/right in the grid — keep important content in the center ~90%. (Buffer guide, same URL; 9to5Mac notes the grid change: “In January [2025], the company started framing photos within rectangles in user profile grids” — same 9to5Mac article.)

Sizes & file limits

• Recommended export: 1080 px wide; 1080×1350 (4:5), 1080×1080 (1:1), 1080×566 (1.91:1), 1080×1440 (3:4). (Buffer; Hootsuite — blog.hootsuite.com/instagram-carousel)
• Image formats: JPG/PNG/BMP/non-animated GIF (Buffer).
• Video slides in carousels: 3–60 seconds, MP4/MOV, max 4 GB (Hootsuite). Carousel videos are not Reels — they don't appear in the Reels tab.
• I could not verify an official image file-size cap from a primary source (Meta doesn't publish one in the pages I could reach; the ads page says “no maximum resolution”). In practice, 1080px-wide JPG exports are far below any limit. Note this is a gap, not a number I'm confident about.

Mixing photos + videos

• Yes — mixed-media carousels are supported natively (photos + videos in one carousel; the 20-slot count is combined). (Sprout FAQ, sproutsocial.com/insights/instagram-carousel; The Verge/MacRumors describe the 20-limit as “photos and videos”.)
• Caveat from Buffer: if you include videos, Instagram can switch the whole carousel to portrait dimensions depending on the sizing mode. For a pure photo series this doesn't matter — keep it photos-only and you're safe.

Post-publish editing (new, 2025–2026)

• You can now reorder a published carousel (long-press + drag) without losing likes/comments — rolled out March 2026, confirmed by Instagram's @creators and Mosseri. You still can't add new slides to an existing post. (PetaPixel, 24 Mar 2026 — petapixel.com/2026/03/24)
• Per-slide captions are rolling out (June 2026): a separate caption for each photo/video in a carousel — perfect for labeling “shot on X-T20” per slide. (Engadget, “Instagram Now Lets You Add A Unique Caption To Each Carousel Slide”, Jun 2026; wersm summary — wersm.com/instagram-tests-captions-on-each-slide-within-carousels). Verify it's live in your app before relying on it.
• Music can be added to photo carousels (rolling out since Apr 2023). (The Verge, Apr 2023, referenced in The Verge's 20-slide article.)
• Collaborative carousels: Instagram tested letting friends add photos to your carousel (Oct 2023); collaborative posts with co-authors are available, and Later (2026) confirms collaborative posts work with carousels. Useful if all 4 friends want the same post on their profiles. (The Verge, 27 Oct 2023 — theverge.com/2023/10/27/23935363; Later — later.com/blog/best-time-to-post-on-instagram)

03 · What the numbers say

Engagement data: carousels vs singles vs Reels (2024–2026)

All major 2024–2026 studies agree: carousels beat single images on engagement, and are the top feed format. Details:

Sources: Socialinsider — socialinsider.io/blog/instagram-benchmarks; Socialinsider carousel guide — socialinsider.io/blog/instagram-carousel; Metricool via PR Newswire — prnewswire.com (also cited by Hootsuite — blog.hootsuite.com/instagram-carousel); Later — later.com/blog/best-types-of-instagram-posts and later.com/blog/schedule-carousel-instagram

What carousels win on

• Reach: Socialinsider's latest data: carousels 3.32% reach rate vs images 3.30% and Reels 3.15%. Adam Mosseri himself has said carousels outperform single photos in reach. (Socialinsider carousel guide.)
• Saves & views: carousels lead in both across every account size — e.g., 10–50K accounts: 8 saves/carousel vs 3/image; 50–100K: 35 vs 10. (Socialinsider benchmarks.) Metricool's 2026 study teaser: carousels get 9× more saves than single-image posts — metricool.com/instagram-research-study
• “Double exposure”: Instagram can re-show a carousel to a viewer who scrolled past it, with a different slide displayed — carousels are the only format that gets a second chance in the feed. (Socialinsider benchmarks; also described by Sprout: “Instagram re-serves carousels to users who didn't engage on the first swipe.”)
• Reels win on raw reach/discovery (Metricool 2024: Reels reach 37.87% of audience) and on comment volume (Socialinsider). For a personal photography series, that's not the goal — carousels are.

Ideal slide count

• Socialinsider: carousels with more than 10 slides get increased reach (their 2026 data). (socialinsider.io/blog/instagram-carousel)
• Counterweight: no reputable public data on swipe-through/drop-off by slide count. The “10–20 slides = more reach” effect is presumably driven by quality longer sets, not padding. Practical stance from Sprout/Hootsuite: “use as many slides as the story demands, then cut. Every frame must earn its place” (Sprout FAQ).
• Swipe-through rates: I could not find a reputable public benchmark for “how many slides the average viewer sees” — treat any specific percentage you see on marketing blogs as unverified. What is documented: the first slide does the heavy lifting (Sprout: “If the first frame doesn't earn attention, the rest of the carousel doesn't get seen”), and because any slide can be the one that gets re-served, every slide should stand on its own (Socialinsider).

04 · How to build the series

Photo-dump / photo-series best practices

• Opening slide = hook. “Start with a strong first slide… a bold visual, a clear promise, or a headline that makes the next swipe feel worth it” (Sprout). For a photo series, the opening slide is your single strongest photo — the “strongest first” plan is correct. Socialinsider: “invest extra time in that very first image… if it sparks curiosity… people are far more likely to swipe.”
• Structure like a story arc. Sprout's framework: “Open with a strong hook, build context slide by slide and close with a takeaway that sticks”; “build momentum as people swipe… save part of the payoff for the end.” For a photo walk: establishing/wide → sequence of details and moments → strongest emotional/closing photo last.
• Variety within consistency. Keep a consistent color grade and consistent ratio across a carousel so it reads as one post (Sprout: “consistent colors, fonts and spacing”; Hootsuite example: Canva “keeps the color palette consistent across slides”). But vary content: wide shots, close-ups, people, details — that's what makes a photo dump feel alive (Later's photo-dump guide: mix of content types is the point — later.com/blog/instagram-photo-dump-trend).
• Text/context slides between photos: the industry norm is minimal for photography series. Text slides are for tutorials/education content (which drives saves); for photos, keep the images clean and put context in the caption — or use the new per-slide captions (rolling out 2026) for one-line notes. If you want an act title (“I · COLOUR”), a single consistent typographic title slide right after the cover is the standard, subtle way (consistent font/color across the 3 posts to tie the series together).
• Cover slide design: single photo, no text overlay, for a personal photography account (matches “lead with a bold visual” — Sprout). Text-overlay covers are a branded-content convention; they read as “brand post”, not personal photo series. To make the 3-post series visually consistent as a set, keep the treatment consistent: same crop style, same grade, covers that echo each other (e.g., each cover a different color story: pink/red for COLOUR, gold/brown for TIME, blue for FORM).
• Closing slide: end strong — either the best “wow” photo or a deliberate coda (the group shot, the scene at night). Sprout: “Your final slide is your strongest conversion opportunity” — for a personal post, the conversion is a comment or a follow, so close with something that rewards the swipe.
• Photo-dump legitimacy: the format is mainstream for personal accounts (“photo dumps” = casual multi-photo posts that convey a story or mood; #photodump has 5M+ posts — Later). Event-recap carousels are an established use (Hootsuite's example list includes event recaps; Later recommends “recap an event” as a photo-dump direction).

05 · Captions and discovery

Caption & discovery (2025/2026 state)

Structure

• Hook line first (front-load: the first ~1–2 lines are what show in the feed before “more” is tapped). Later's social team: start with a one-liner that stops the scroll; then the body; then a clear CTA on its own line (line breaks matter — Later deliberately separates the CTA with a line break); end by encouraging conversation (a question). (later.com/blog/the-ultimate-guide-to-writing-good-instagram-captions)
• Long-form captions are fine and are keyword real estate (Hootsuite SEO guide), but only if they stay engaging.

Search / SEO

• Captions ARE searchable; comments are NOT. Instagram's official search ranking factors (per Meta's “how Instagram search works” explainer): search text (usernames, bios, captions, hashtags, locations) → user activity → popularity signals → recency. So put keywords in the caption, not the first comment. (Hootsuite — blog.hootsuite.com/instagram-seo)
• So “Malang”, “Jodipan”, “Kayutangan”, “Kampung Warna-Warni”, “photo walk”, “street photography” should appear naturally in the caption text — people search those.
• Use the location tag on each post (locations are a search surface — same Hootsuite source).

Hashtags — big change in Dec 2025

• Instagram capped hashtags at 5 per post/reel (previously 30), announced Dec 2025; Mosseri called the “more hashtags = more reach” belief a “hashtag myth” — hashtags now primarily organize/search, and AI recommendations drive discovery. (Times of India, 19 Dec 2025 — timesofindia.indiatimes.com; The Verge, 18 Dec 2025 “Instagram wants to limit hashtag spam”; Social Media Today, 18 Dec 2025.)
• Practical 2026 rule: 3–5 highly relevant hashtags, camelCase them for accessibility, put them in the caption (not comments). (Hootsuite SEO guide; Mosseri's own earlier guidance was 3–5.)
• For these posts: something like #Malang #Jodipan #Kayutangan #StreetPhotography #PhotoWalk per post, rotated to match the act (e.g., #BlueHour for Act III, #Heritage for Act II).

Alt text

• Add alt text per slide (each carousel photo gets its own alt-text slot, editable even after posting; max 100 characters, separate from the caption; used by screen readers and by Instagram's algorithm to understand the image). Write descriptive, not keyword-stuffed: e.g., “Painted houses and stairways in Kampung Warna-Warni Jodipan, Malang, at midday.” (Hootsuite — blog.hootsuite.com/instagram-alt-text)

06 · When and how often

Posting strategy

Best time to post — Indonesia

• There is no reliable Indonesia-specific public dataset for Instagram posting times. Indonesian media publish “best time” articles (e.g., CNN Indonesia, “Kapan Waktu Terbaik Buat Posting di Instagram?”), but I could not verify a primary study behind them, and the numbers they repeat are typically recycled global data. Treat any “best time for Indonesia” claim as unverified.
• Closest reputable data (global, audience-localized):
  • Later (6M+ posts, Jan–Oct 2024): best overall = 5 AM in your audience's local time zone; for carousels specifically: Tuesday at 5 AM (with 3–6 AM consistently above average). Data is normalized to followers' local time. (later.com/blog/best-time-to-post-on-instagram)
  • Metricool 2024: most users online globally 8 PM, ideal publishing window 7–9 PM, best days Wednesday & Friday. (PR Newswire release, same study as §03.)
• The two disagree (5 AM vs 7–9 PM) — which is exactly why the correct move is: check Instagram Professional Dashboard → “Most Active Times” for your followers, and use the studies only as starting hypotheses. (Later's article recommends the same.)
• For a WIB audience, Later's model would point to ~05:00 WIB (first-scroll window, low competition); Metricool's to ~19:00–21:00 WIB (evening browsing). Test one, then the other.

Post all 3 carousels at once, or spaced?

• No data directly compares “3 posts same day” vs “1/day” — I won't invent numbers.
• Standard practice from the guides above: each post needs its own window of early engagement (the algorithm weighs fast initial interaction — Later best-time article; Hootsuite SEO article) and posting a burst dilutes that. Recommendation: one per day (or at minimum, hours apart), ordered COLOUR → TIME → FORM (matching the day's chronology, and each cover echoing the next). Consistency of the series is carried by the visuals, not by posting together.
• The day itself is 16 Aug 2026 (eve of Indonesian Independence Day — flags and crowds at Tugu are part of the story; the plan doc already notes this). If there's any time-sensitivity (e.g., “we did this yesterday”), post Act I within 24–48h of the event; the series can then continue on consecutive days.

“Shoot for the grid” (aspect-ratio planning while shooting)

• Sensors here are 3:2 (Fuji X-T20, Sony a6100/a6300) or 4:3 (Xiaomi 14T); Instagram's recommended feed ratio is 4:5 — meaning a 3:2 frame loses a lot of width when cropped to 4:5. Practical consequence for the shoot day: compose with the vertical crop in mind — leave headroom above/below your subject, keep key content in the central ~90% of the frame, and don't put critical elements at the far left/right edges.
• If grid aesthetics matter, note the profile grid now shows 3:4 crops (Buffer) — slightly taller than 4:5. Simplest robust approach: crop everything to 4:5 in post (or 3:4 if you want zero grid cropping), export at 1080×1350 (or 1080×1440), and keep the same crop on all slides of all 3 posts.
• (Buffer: “keep important visuals inside that 3:4 safe zone” — buffer.com/resources/instagram-image-size)

07 · The comparison post

Same-scene / camera-comparison post — existing conventions

There is no single canonical “Instagram camera comparison” template, but there are well-established conventions from photography media and community practice:

• The blind-test / A-B convention (the strongest one). Photography publications present same-scene camera comparisons as labeled side-by-side images with the reveal at the end so viewers judge on merit, not brand bias. PetaPixel's Fujifilm-vs-Canon color-science blind test is the canonical example: same scene, multiple cameras, presented as unlabeled options, “answer key” at the end, and the audience is told to pick the image that sparks the best emotional response rather than pixel-peep. (PetaPixel, 7 Oct 2019 — petapixel.com/2019/10/07) This maps directly onto a carousel: one slide per camera, unlabeled or weakly labeled, with the camera names revealed in the final slide/caption.
• The label-every-slide convention. The default way multi-camera takes are shown in carousels is consecutive slides with a label per slide (“shot on X-T20 / a6100 / a6300”). The new per-slide captions (rolling out 2026, §02) make this clean — label each slide with its camera without polluting the main caption. If per-slide captions aren't available in the app yet, a small consistent text label on each slide (same font, same position) is the standard workaround.
• Pair/side-by-side crops. For detail comparisons (sharpness, noise, color), the convention is a single composite slide: two 100% crops of the same area side by side, labeled (this is how PetaPixel/Fstoppers-style reviews do it; the blind-test article above shows the side-by-side layout). For your “same scene from the bridge at 14:00” set, you could do both: full-frame slides per camera plus one composite crop slide.
• Same-scene rules of thumb from testing practice: match focal length/field of view as closely as possible, same exposure intent (don't equalize in post — the differences are the point), and present RAW/JPEG honestly. (PetaPixel blind-test article: shots taken with native lenses, presented as RAW + JPEG; viewers told not to compare depth of field, only response.)
• Engagement hook: turn it into a question (“A, B, C — which one?”) — asking a question in the caption is the standard CTA for this kind of post (Later's caption guide: conversation-inducing captions; Sprout: comparison/choice carousels like Chipotle's “choose your trio” drive comments).
• Alternative from brand practice: “before-and-after”-style carousels are a proven format (Hootsuite/Later both list them); a camera comparison is conceptually the same “reveal” mechanic — tease first, reveal last.
• Since the group's plan already includes the 14:00 bridge shot from all cameras, the 4th post practically writes itself: 3–4 slides (one per camera) + 1 composite crop slide + caption asking “which look do you prefer?” — and optionally a collaborative post so all friends share it (The Verge, Oct 2023, §02).

Sources

Where these came from

All accessed August 2026.

• The Verge — “You can now share 20 slides in one Instagram post” (8 Aug 2024): theverge.com/2024/8/8/24216111
• MacRumors — “Instagram Now Allows Up to 20 Photos and Videos for Carousel Posts” (8 Aug 2024): macrumors.com/2024/08/08/instagram-20-photos-carousel-posts
• Metricool — “Instagram Carousel Limit Increased to 20 Photos or Videos” (9 Aug 2024): metricool.com/instagram-increases-carousel-content-limit
• Sprout Social — “A guide to crafting seamless Instagram carousel posts” (2026 ed.): sproutsocial.com/insights/instagram-carousel
• Buffer — “Instagram Post Size Guide 2026” (incl. 3:4 grid, carousel crop rules): buffer.com/resources/instagram-image-size
• Hootsuite — “How to make the most of Instagram carousels in 2025” (specs, video limits, Metricool citation): blog.hootsuite.com/instagram-carousel
• 9to5Mac — “No more cropping: Instagram changes standard photo aspect ratio” (29 May 2025): 9to5mac.com/2025/05/29/instagram-changes-standard-photo-aspect-ratio
• Meta Business Help Center — “Design Requirements for Instagram Feed Ads” (feed ratio range 1.91:1–9:16, 4:5 recommended, ≥1080 px): facebook.com/business/help/430958953753149
• Socialinsider — “2026 Instagram Organic Engagement Benchmarks” (35M posts): socialinsider.io/blog/instagram-benchmarks
• Socialinsider — “Instagram Carousel Post: How to Use It for Your Brand” (reach by format, >10 slides): socialinsider.io/blog/instagram-carousel
• Later — “Best Instagram Posts Types: Carousel, Image & Video Ranked” (81M posts): later.com/blog/best-types-of-instagram-posts
• Later — “Instagram Carousel Post: A Guide for Creators” (+19% vs single images): later.com/blog/schedule-carousel-instagram
• Later — “Best time to post on Instagram in 2026” (6M posts; carousels Tuesday 5 AM): later.com/blog/best-time-to-post-on-instagram
• Metricool 2024 Instagram Benchmark Report (PR Newswire, 9 Sep 2024): prnewswire.com
• Metricool — 2026 Instagram Study page (9× saves claim): metricool.com/instagram-research-study
• Later — “Photo Dumps: What They Are and How to Start Posting Them”: later.com/blog/instagram-photo-dump-trend
• Later — “8 Ways to Make Instagram Carousel Posts Your Audience Loves”: later.com/blog/instagram-carousel-posts
• Later — “How to Write Instagram Captions”: later.com/blog/the-ultimate-guide-to-writing-good-instagram-captions
• Hootsuite — “Instagram SEO made easy” (search ranking factors, captions searchable, 3–5 hashtags): blog.hootsuite.com/instagram-seo
• Hootsuite — “Why Instagram Alt Text Matters” (100-char limit, per-slide alt text): blog.hootsuite.com/instagram-alt-text
• Times of India — “After Instagram head Adam Mosseri says hashtags are myths, app limits hashtag count on posts” (19 Dec 2025): timesofindia.indiatimes.com
• The Verge — “Instagram tests collaborative carousels your friends can add to” (27 Oct 2023): theverge.com/2023/10/27/23935363
• PetaPixel — “Photographers Can Now Reorder Their Pictures on an Instagram Carousel Post” (24 Mar 2026): petapixel.com/2026/03/24
• Engadget — “Instagram Now Lets You Add A Unique Caption To Each Carousel Slide” (Jun 2026); summary: wersm.com/instagram-tests-captions-on-each-slide-within-carousels
• PetaPixel — “Fujifilm vs Canon Color Science Blind Test” (7 Oct 2019): petapixel.com/2019/10/07/fujifilm-vs-canon-color-science-blind-test-can-you-tell-the-difference

Honest gaps (things I could not verify)

• No reliable public data on swipe-through rates (average slides viewed) — any percentages you see on marketing blogs are unverified.
• No reliable Indonesia-specific best-posting-time data — use Later/Metricool global numbers as hypotheses and your Professional Dashboard as truth.
• No official image file-size cap found in primary sources (Meta's pages publish “no maximum resolution”; only video has a documented cap: 4 GB via Hootsuite). 1080px JPG exports are never near limits.
• help.instagram.com itself returned errors when fetched directly (bot protection); official-spec claims were cross-verified via Meta's Business Help pages and multiple independent 2025–2026 tool guides instead.